[jira] [Updated] (CASSANDRA-9241) ByteBuffer.array() without ByteBuffer.arrayOffset() + ByteBuffer.position() is a bug

["Ariel Weisberg (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/CASSANDRA-9241?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Ariel Weisberg updated CASSANDRA-9241:
--------------------------------------
    Component/s: Testing

> ByteBuffer.array() without ByteBuffer.arrayOffset() + ByteBuffer.position() is a bug
> ------------------------------------------------------------------------------------
>
>                 Key: CASSANDRA-9241
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-9241
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Testing
>            Reporter: Ariel Weisberg
>            Assignee: Ariel Weisberg
>            Priority: Minor
>             Fix For: 3.2
>
>
> I found one instance of this on OHCProvider so it make sense to review all usages since there aren't that many.
> Some suspect things:
> https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/utils/FastByteOperations.java#L197
> https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/db/ColumnFamilyStore.java#L1877
> https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/gms/TokenSerializer.java#L40
> https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/io/compress/CompressedRandomAccessReader.java#L178
> https://github.com/apache/cassandra/blob/trunk/tools/stress/src/org/apache/cassandra/stress/operations/predefined/CqlOperation.java#L104
> https://github.com/apache/cassandra/blob/trunk/tools/stress/src/org/apache/cassandra/stress/operations/predefined/CqlOperation.java#L543
> https://github.com/apache/cassandra/blob/trunk/tools/stress/src/org/apache/cassandra/stress/operations/predefined/CqlOperation.java#L563
> I made this list off of 8099 so I might have missed some instances on trunk. FastByteOperations makes me cross eyed so it is worth a second pass to make sure offsets in byte buffers are handled correctly.
> Generally I like to use the full incantation even when I have done things like allocate the buffer on the stack locally for copy pasta/refactoring reasons and to make clear to new users how the API is supposed to work.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)